Image sensor based steering signal for a digital actuator system

Abstract

Autonomous driving is an emerging field of research. The related industry is one of the most expensive areas nowadays. The core of these complex controlling systems is the perception of the environment and the usage of actuators for changing supported by sensors to give obvious feedback about the change of state. Steering controlling is such a subsystem. In the real-sized modern car, there are several methods for implementing feedback loops for it, such as torque and angle sensors. In this paper, we concentrate on extending our model-based research and development autonomous vehicle platform – DAVE to be able to study the hard conditions safely and cost-effectively. This work presents a method for a new sensor signal integrated into our CAN-BUS system to give feedback about the steering movement of the wheels to a digital steering controller using a rear-view camera. The advantage of using a rear-view blind-spot camera is that it is already in place, and no additional hardware is needed to use it as a pseudo angle sensor.